Note: Descriptions are shown in the official language in which they were submitted.
CA 02318107 2000-07-20
WO 99/38137 PCT/US99/01415
UNIFORM MOTIVATION FOR MULTIPLE COMPUTERP-ASSISTED TRAINING
SYSTEMS
SPECIFICATION
CROSS REFERENCE TO RELATED APPLICATIONS
This application is related to co-pending ( 1 ) U. S. Patent Application
Serial No.
(Atty Docket No.: P-2076/702C2), entitled "Remote Computer-Assisted
Compliance Monitoring System," filed December 22, 1997 (hereinafter "the
Compliance
Monitoring Application"), {2) U.S. Patent Application Serial No. (Atty Docket
No.: P-2071/702C1), entitled "Migration Mechanism for User Data from One
Client
Computer System to Another," filed December 22, 1997 (hereinafter "the
Migration
Application"), and (3) U. S. Patent Application S/N 08/ - filed , ~ 1997 by
William M. Jenkins et al. and entitled "Method and Device for Training of
Sensory
Perceptual System in LLI Subjects" (hereinafter "the Training Application")
which are all
incorporated herein by reference in their entirety for all purposes.
FIELD OF THE INVENTION
The present invention relates to computer-assisted training systems and, in
particular, to a particularly effective mechanism by which a user of the
training system is
motivated to continue with a training program.
BACKGROUND OF THE INVENTION
For some years now, many attempts have been made to harness the popularity and
nearly addictive properties of computer and video training exercises for the
purpose of
training and education. Since arcade and home style video training exercises
are generally
controlled by one form or another of an electronic computer, all such training
exercises,
including those played using a general purpose computer, are referred to
herein as
computer training exercises. Computer training exercises have the advantage
that the
challenge itself is interesting to the user. For example, if the user is
challenged to use user
CA 02318107 2000-07-20
WO 99/38137 PCT/US99/01415
-2
input devices to shoot and kill space aliens, tfe act of shooting and killing
the space aliens
is, for certain users, fun per se. In general, computer training exercises
present a simulated
threat to the user and give the user a simulated defense or weapon to overcome
the
simulated threat. Such tends to challenge the competitive spirit of the user
and motivate
the user to overcome the simulated threat.
Other computer training exercises simulate a direct competition between the
user
and a simulated or actual opponent. Such training exercises include racing and
fighting
training exercises in which the user respectively controls a vehicle to race
an opponent's
vehicle through a race course or controls a fighting character to fight an
opposing fighting
character using user input devices. The opposing vehicle or fighting character
can be
controlled by the computer or by another user. These training exercises also
challenge the
competitive spirit of the user and motivate the user to out-perform the
simulated or actual
opponent.
Training programs typically include repetitive exercises to improve certain
skills or
cognitive abilities of the user. Some computer-based training programs use the
simulated
threat paradigm of computer training exercises to motivate the user to perform
such
repetitive training exercises. For example, the PalmPilot digital personal
organizer
available from US Robotics, Inc. of Skokie, Illinois, includes a training
process which
teaches the user to use a special alphabet that the PalmPilot digital personal
organizer can
recognize through a touch-sensitive pad. In this training exercise, letters
and numbers
move down the screen, apparently at the user, to form a simulated threat. In
response, the
user can simulate destruction of the threatening letters and numbers by
writing the
threatening letters and numbers using the touch-sensitive pad. By presenting
the user with
a simulated threat and incorporating a training exercise into a simulated
defense or weapon,
the user can be motivated to perform the training exercise repeatedly. Such a
training
program can be relatively effective if the training exercise is readily
adaptable to a
simulated defense or weapon and the skills improved by the training exercise
are relatively
simple.
Certain training exercises do not lend themselves as readily to the
threat/defense
training exercise paradigm described above. For example, some training
exercises are not
so readily adapted to a simulated defense or weapon. An example of such a
training
CA 02318107 2000-07-20
WO 99/38137 - 3 - PGTNS99/01415
exercise is the recognition of a language phoneme by the user, e.g.,
distinguishing "shu"
from "chu." In addition, some training programs are so long, e.g., more than
one hour per
day for several weeks, that even the threat/defense training exercise paradigm
loses its
motivation efficacy.
A particularly helpful computer-based training program is described in the
above-
referenced patent applications and further in the Training Application. The
described
training program involves student participation for more than one hour per
day, at least five
(5) days per week, for eight (8) weeks. Such a program is long and maintaining
interest
and motivation in the various training exercises over such a long program is
of particular
importance. Exacerbating the problem is the fact that the users of the
described computer-
based training program are children. In addition, the objects of the training
exercise are
language-oriented in nature and therefore do not lend themselves to simulated
weapons
actuation as compared to relatively simple motor skills, e.g., typing, which
are more
relatively readily adaptable to simulated weapons actuation. What is therefore
needed is a
system for maintaining interest and motivation in the training exercises over
the entire
training program.
SUMMARY OF THE INVENT_rON
In accordance with the present invention, motivation mechanisms which are
independent of stimuli of training exercises to which the user is to respond
motivate the
user to respond correctly and quickly to the stimuli. The stimuli of the
training exercises
are specifically selected and designed to challenge and improve a cognitive
ability of the
user. As a result, the manner in which the stimuli can be adapted to create
and maintain
interest of the user in the training exercises is significantly limited.
Therefore, the
motivation mechanisms of the training exercises therefore take on added
importance.
A standardized token economy is used to motivate the user to respond correctly
and quickly to stimuli of the training exercises. Points are awarded for
correctly
responding to stimuli and are the currency of the token economy. To add
motivation, the
points can be used as a currency to exchange for physical rewards such as
toys, T-shirts,
and even tickets to sporting events. The token economy is standardized in that
the amount
CA 02318107 2000-07-20
WO 99/38137 - 4 - PCTNS99/01415
of points accumulated during use of one training exercise for a particular
amount of time
with a particular degree of proficiency is approximately equal to the amount
of points
accumulated during use of another training exercise for the same amount of
time with the
same degree of proficiency. Therefore, training exercises with particularly
quick
stimuli/response exchanges are not favored as "point friendly" at the expense
of training
exercises in which time required to present stimuli and receive corresponding
responses
take longer. In this way, preference for particular training exercises based
upon the
quickness with which a user can accumulate points is avoided.
In addition, while the stimuli of the training exercises generally remain
unchanged in
substance and change only in terms of levels of difficulty, milestones of
achievement are
noted by changes in the context within which the training exercises are
presented to the
user. Such changes in context include changes in one or more characteristics
of a scene
within which the training exercise is displayed to the user. Each change of
context is
brought to attention with fanfare using a context change animation. For
example, a janitor
character is shown to paint a portion of the scene a different color in such a
context change
animation.
Progress indicators represent to the user progress toward a short term goal.
Each
correct response by the user to stimuli of a training exercise moves the
progress indicator
closer to a position indicating achievement of the short term goal. Each move
of the
progress indicator is smooth to give the appearance of progression rather than
a discrete
change in state. Such a move is in the form of a progress animation which is
entertaining
and which provides motivation to the user, both in terms of the entertainment
value of the
progress animation and in terms of the progression toward the short term goal.
These motivation mechanisms provide the user with significant motivation
notwithstanding limitations in the nature and adaptation of stimuli presented
to the user by
various training exercises.
BRIEF DESCRIPTION OF THE DRAWINGS
Figures 1-80 are screen views of computer-implemented training exercises in
accordance with the present invention.
CA 02318107 2000-07-20
WO 99/38137 PCT/US99/01415
-S
Figure 81 is a block diagram of a computer system in which training exercises
according to the present invention execute.
Figure 82 is a block diagram of a training exercise of Figure 81.
Figure 83 is a logic flow diagram of the progress indication and bonus point
determination of a training exercise in accordance with the present invention.
DETAILED DESCRrn~rr~rr
In accordance with the present invention, a number of training exercises use
motivation mechanisms independent of the training stimuli presented to the
user to
motivate the user to continue with extended use of the training exercises and
to user the
training exercises to the full extent of the user. The stimuli of each of the
training exercises
is specifically selected to challenge and improve a cognitive ability of the
user and therefore
is limited in the ways the stimuli can be changed to create and maintain user
interest in the
tr&ining exercises. Accordingly, the motivation mechanisms of the present
invention are of
particular importance in such training exercises.
The Tag xerci t
The motivation systems of the present invention are implemented, in this
illustrative
embodiment, in a training process which includes a number of training
exercises designed
to improve the language comprehension of language learning impaired (LLI)
individuals.
Appreciation of the motivation systems according to the present invention is
facilitated by a
brief discussion of each of the training exercises of the training process. In
each of the
training exercises described below, the stimuli presented to the user are
designed primarily
to improve a cognitive ability of the user, specifically, the ability to
perceive and understand
spoken language.
The requirement that the stimuli challenge and improve a cognitive ability of
the
user inherently restricts the stimuli. As a result, adaptation of the stimuli
themselves during
extended use of the training exercises by the user to help maintain interest
of the user in the
training exercises is very restricted. In general, the stimuli are adapted
only to fiuther
challenge and improve the cognitive ability of the user as the user's
cognitive abilities
improve. However, the stimuli appear the same to the user over time and the
user can
CA 02318107 2000-07-20
WO 99/38137 - 6 - PCTNS99/01415
become quite bored with the stimuli and may approach the stimuli with
ambivalence and
indifference.
By comparison, the primacy purpose of any stimuli preseccted to a user of a
computer training exercise is entertainment. Any improvement in a cognitive
ability of the
user, perhaps hand-eye coordination in control simulated defenses and/or
weapons using
user input devices, is merely incidental. In addition, the stimuli of computer
training
exercises, in the form of simulated threats or opponents, can change
dramatically over time
to maintain the user's interest in the computer training exercises. For
example, space
aliens, which are unarmed at lowers levels of play, can be given simulated
laser-firing
weapons at higher levels to further challenge and entertain a user. Thus, the
stimuli of
computer training exercises can change considerably to maintain user interest
in such
computer training exercises since the stimuli are not rest~cted to improving a
particular
cognitive ability of the user.
Old MacDonald's Flaring Farm
Figures 1-16 are screen views of a training exercise called "Old MacDonald's
Flying Farm" which is used to improve language comprehension of LLI users. Old
MacDonald's Flying Farm is a limited hold reaction time procedure test in
which a flying
object 102 (Figure 1) flies around until the user grabs flying object 102
using a cursor 104.
Flying object 102 is shown as so grabbed in Figure 2. The user grabs flying
object 102
with cursor i 04 using conventional drag-and-drop user interface techniques
involving
physical manipulation of a user input device in a manner described more
completely below.
The user holds flying object 102 while a phoneme is repeatedly played for the
user.
For example, the audible sound, "si," can be repeated through a computer sound
card and
loudspeaker as described more completely below. At some point, a similar but
distinct
phoneme, e.g., "sti," is substituted for the repeated phoneme. The user is
expected to
recognize the substituted phoneme as distinct and so indicate by releasing
flying object 102.
If the user releases flying object 102 prior to substitution .of the distinct
phoneme or fails to
release flying object 102 within a predetermined period of time following
substitution of the
distinct phoneme, the user's response is characterized as incorrect, and
flying object 102 is
shown, through computer animation, to fly away as shown in Figure 3.
Conversely, if the
CA 02318107 2000-07-20
WO 99/38137 - 7 - PCT/US99/01415
user releases flying object 102 within the predetermined period of time
following
substitution of the distinct phoneme, the user's response is characterized as
correct and
flying object 102 is shown, through computer animation, to fly into a home 110
as shown
in Figure 4.
Thus, Old MacDonald's Flying Farm trains the user to recognize distinctions
between similar but distinct phonemes, such as (i) "si" and "sti" and (ii)
"chu" and "shu."
The stimuli of Old MacDonald's Flying Farm, i.e., the similar but distinct
phonemes, are
selected to challenge, and therefore improve, a cognitive ability of the user,
namely, to
perceive subtle distinctions between similar but distinct phonemes. The
phonemes
themselves have no real entertainment value in and of themselves for the user.
Instead,
emertainment value is built into the context surrounding the stimuli, namely,
an animated
flying farm animal which many users find to be humorous. In addition, a sense
of order
motivates the user to coax the flying farm animal into a home, e.g., home 110
which is a
barn in which the cows belongs from the perception of the user. Similarly, the
flying chick
of flying object 202 (Figure 15) belongs under a mother hen, e.g., home 210.
(', irc~~s Sequence
Figures 17-34 are screen views of a training exercise called "Circus Sequence"
which is used to improve language comprehension of LLI users. Circus Sequence
trains a
user to distinguish frequency sweeps, i.e., audible sounds with simple tones
sweeping
continuously through a range of frequencies. In particular, the user is
trained to distinguish
between two frequency sweeps covering the same frequency spectrum, one
sweeping from
low to high frequency and the other sweeping from high to low frequency. The
former is
called an up sweep and the latter is called a down sweep. The duration of the
frequency
sweeps and the inter-stimuli interval (ISI), i.e., pause between frequency
sweeps, are
adjusted according to the ability of the user to distinguish the frequency
sweeps as
measured in previous play of Circus Sequence by the user.
In Circus Sequence, the user indicates readiness to hear the frequency sweeps
by
clicking on a button 302 (Figure 17). The user clicks button 302 using
conventional user
interface techniques by physical manipulation of user input devices as
described more
completely below. A button 316A (Figure 18) is associated with an up sweep and
another
CA 02318107 2000-07-20
WO 99/38137 PCT/US99/01415
-g
button 316B is associated with a down sweep. Two sweeps are played to the user
and the
user identifies the sweeps by pressing associated ones of buttons 316A-B. For
example, if
the user hears two up sweeps, the user presses button 316A twice. If the user
hears a
down sweep followed by an up sweep, the user presses button 316B then presses
button
316A to correctly identify the sweeps. Accordingly, Circus Sequence trains the
user to
distinguish up sweeps and down sweeps.
The stimuli of Circus Sequence, i.e., the up sweeps and down sweeps, are
selected
to challenge, and therefore improve, a cognitive ability of the user, in
particular, the ability
of the user to recognize quick changes in frequency which are common in spoken
language.
The frequency sweeps have little, if any, inherent entertainment value for the
user.
Accordingly, extrinsic motivation must generally be provided to encourage the
user to
listen attentively to the up and down sweeps and to correctly identify them.
Such
motivation is described more completely below.
Block Commander
Figures 35-40 are screen views of a training exercise called "Block Commander"
which is used to improve language comprehension of LLI users. Block Commander
improves a user's language comprehension. Specifically, Block Commander
audibly directs
a user to perform an action involving one or more of a number of objects 404
(Figure 35)
using a cursor 402. Using a computer-synthesized voice played through a sound
card and
loudspeaker, for example, Block Commander can direct the user to "touch the
blue
square." Using conventional drag-and-drop user interface techniques involving
physical
manipulation of user input devices, the user complies with the direction given
by Block
Commander. Compliance with the direction indicaxes that the user successfully
understood
the direction. Any act by the user other than the act directed by Block
Commander
indicates that the user failed to properly understand the direction.
The computer-synthesized voice used by Block Commander to give audible
directions to the user is processed and modified in a manner described more
completely in
the Training Application and that description is incorporated herein by
reference. The
modification of the voice makes the voice more easily understood by LLI
individuals. The
degree to which the computer-synthesized voice is modified is adjusted
according to the
CA 02318107 2000-07-20
WO 99/38137 - 9 - PCTNS99/01415
language comprehension ability of the user a~ measured in previous play of
Block
Commander. Thus, as the language comprehension ability of the user improves,
the degree
to which the computer-synthesized voice is modified is reduced to thereby
challenge the
language comprehension abilities of the user. Accordingly, Block Commander
trains and
improves language comprehension of the user.
The stimuli ofBlock Commander, i.e., the audible directions, are selected to
challenge and therefore improve the language comprehension abilities of the
user - a
cognitive ability of the user. For example, the direction, "touch the small
blue square,"
requires that the user distinguish "small" from "large," "blue" from other
colors, and
distinguish "square" from "circle." Other complexities of speech can be
included in the
stimuli as the language comprehension of.the user improves. For example, the
user can be
directed as follows: "Except for the blue one, touch the circles." The user is
therefore
challenged to understand exclusion. In addition, the user can be directed as
follows:
"Before touching the red circle, touch the blue square." Such challenges the
user to
understand compound phrases and sequencing. Similarly, the user can be
directed as
follows: "Place the green square between the white circle and the red circle."
These stimuli are specifically designed to challenge, and therefore improve,
specific
language comprehension abilities of the user and have little or no inherent
entertainment
value in and of themselves. Accordingly, motivation to the user to comply with
the audible
directions must generally be provided independently of the audible directions
themselves.
Phonic Match
Figures 41-47 are screen views of a training exercise called "Phonic Match"
which
is used to improve language comprehension of LLI users. Phonic Match
associates audible
words with respective ones of panes 504 (Figure 41). The audible word
associated with a
given one of panes 504 is played for the user in response to clicking on the
pane with a
cursor 502. Each audible word is associated with two paired ones of panes 504
and the
user is to recognize and identify both panes of each pair. The user identifies
the panes of a
pair by clicking on the paired panes in succession. For example, clicking on
one of panes
504 the user hears "tag" and clicking on a second one of panes 504 the user
hears "bag."
The user has not identified a matching pair of panes 504. Conversely, if the
user clicks on
CA 02318107 2000-07-20
WO 99/38137 - 1 O - PCTNS99/01415
two of panes 504 and hears "tag" both times, the user has identified a
matching pair. When
a matching pair of panes 504 is recognized and identified by the user, those
panes are
removed from the display as shown in Figure 42.
Phonic Match challenges and improves the user's ability to distinguish
similarly
sounding words and the user's memory capacity simultaneously. The audible
words are
modified in accordance with previously determined ability of the user to
comprehend
speech in the manner described above. In addition, the number of panes 504 can
be
increased in accordance with the user's previous performance of Phonic Match.
For
example, Figure 41 shows four (4) panes 504. Figure 45 shows nine (9) panes
504B.
Figure 46 shows sixteen (16) panes 504C with two matched panes removed after
identification by the user in the manner described above.
The stimuli, namely the audible words, are specifically selected to challenge
and
improve a cognitive ability of the user, i.e., to remember and distinguish
similarly sounding
words. For example, the audible words can include "tag," "tack," "tab," "bag,"
"back,"
"pack," "pick," and "tick." Since these audible words are specifically
selected to challenge
and improve the user's ability to remember and distinguish similarly sounding
words, these
words have little or no inherent entertainment value. As a result, motivation
independent
of the stimuli themselves is provide as described more completely below to
encourage the
user to strive to remember and correctly identify matching phonemes.
Figures 48-58 are screen views of a training exercise called "Phonic Word"
which is
used to improve language comprehension of LLI users. Phonic Word tests and
trains a
user's ability to distinguish similarly sounding words. In general, Phonic
Word shows the
user two images 602A B (Figure 48) and audibly directs the user to identify
one of images
602A-B. For example, Phonic Word can play a computer-synthesized voice which
tells the
user to "show me thumb." Using a cursor 612, the user identifies one of images
602A-B.
If the user identifies image 602A, the user correctly identifies the image of
the thumb and
indicates a correct distinction between "thumb" and "tongue." Conversely, if
the user
identifies image 602B, the user has failed to distinguish "thumb" from
"tongue" and so
indicates to Phonic Word.
CA 02318107 2000-07-20
WO 99/38137 - 11 - PC'T/US99/01415
The computer-synthesized voice is modified in accordance with previously
determined ability of the user to distinguish similarly sounding words in the
manner
described above. As the user's ability to distinguish similarly sounding words
improves,
the degree to which the computer synthesized voice is modified is decreased to
further
challenge the user. Accordingly, Phonic Word trains the user to distinguish
similarly
sounding words.
The stimuli, namely the similarly sounding words, are specifically chosen to
challenge and improve the user's ability to distinguish similarly sounding
words and
generally not for entertainment of the user. As a result, the similarly
sounding words have
little, if any, entertainment value for the student. Motivation to listen
attentively and to
correctly comply with the audible direction to identify one of images 602A-B
is provided
independently from the audible directions themselves as described more
completely below.
Phoneme Identification
Figures 59-69 are screen views of a training exercise called "Phoneme
Identification" which is used to improve language comprehension of LLI users.
Phoneme
Idemification tests and improves the user's ability to select and identify a
particular
phoneme from two similarly sounding phonemes. First, Phoneme Identification
plays a
target phoneme which the user is to identify. Next, Phoneme Identification
plays two
similarly sounding phonemes, one of which is the target phoneme.
To provide the user with a mechanism by which to identify each of the
similarly
sounding phonemes, Phoneme Identification associates each of the phonemes with
a
graphical image, i.e., each of images 810A-B (Figure 67). Phoneme
Identification animates
images 810A B to give the appearance to the user that each of images 810A B
speaks a.
respective one of the similarly sounding phonemes. Figure 67 shows image 810A
animated
in the simulated act of speaking the first of the similarly sounding phonemes.
Figure 68
shows image 810B animated in the simulated act of speaking the second of the
similarly
sounding phonemes.
The user identifies the first or second of the similarly sounding phonemes as
the
target phoneme by clicking on image 810A or image 810B, respectively, with
cursor 804
by physical manipulation of user input devices in conventional user interface
techniques.
CA 02318107 2000-07-20
WO 99/38137 - 12 - PCT/US99/01415
The first, second, and target phonemes can bs modified in the manner described
above to
be more easily recognizable to a person with LLI. The degree of phoneme
modification, as
well as the duration of an ISI between presentation of the first and second
phonemes, can
be adjusted according to the ability of the user to recognize target phonemes
among
similarly sounding phonemes as measured by the user's prior performances
during playing
Phoneme Identification.
In this way, Phoneme Identification tests and trains the user's ability to
recognize
particular phonemes from among similarly sounding phonemes. The stimuli, i.e.,
the
similarly sounding phonemes, are selected specifically to challenge and
improve the user's
ability to remember and distinguish similarly sounding phonemes and not to
entertain the
user. As a result, the stimuli generally have no inherent entertainment value
for the user.
Motivation for the user to listen attentively to the stimuli and to correctly
identify the target
phoneme is therefore provided independently of the stimuli themselves as
described more
completely below.
i~g3tage Comorehen~~on Builder
Figures 70-80 are screen views of a training exercise called "Language
Comprehension Builder" ("LCB") which is used to improve language comprehension
of
LLI users. LCB tests and trains the user's ability to understand subtle
distinctions in
spoken language. Specifically, LCB presents to the user between a number of
images
902A D and audibly directs the user to identify one of the images. The images
are selected
so as to be similar to one another and rely upon subtle language distinctions
to specifically
request the user to identify only one of the images. For example, image 902A
shows a car
that is broken, image 902D shows a cup that is broken, and image 902C shows a
cup.
LCB plays an audible direction to the user to "show me the cup is broken." To
properly
identify image 902D, the user has to understand the distinction between "cup"
and "car"
and further has to comprehend "broken."
The user identifies one of images 902A-D by clicking on one of images 902A-D
through control of a cursor 904 using conventional user interface techniques
and physical
manipulation of user input devices. Correctly identifying the one of images
902A D
indicates that the user correctly understood the audible direction given by
LCB. Identifying
CA 02318107 2000-07-20
WO 99/38137 - 13 _ PCT/US99/01415
one of images 902A-D other than the image i3entified audibly by LCB indicates
that the
user failed to understand the audible direction. The audible direction can be
modified in the
manner described above to be more easily understood by user's with LLI. The
degree of
such modification, as well and the speed with which the oral direction is
given, can be
adjusted according to the language comprehension skills of the user as
measured by
previous play sessions with LCB by the same user.
Thus, LCB tests and trains the user's ability to understand subtle
distinctions in
spoken language. The stimuli, which are the audible directions in this case,
are specifically
chosen to challenge and improve a cognitive ability of the user, namely, to
comprehend
subtle distinctions and constructs of spoken language. In one embodiment, the
audible,
audible directions are selected from the known, standardized, C~rtis_Yumada
Comprehensive Language.Examination (CYCLE) language assessment test. The
audible
directions have little, if any, inherent entertainment value for the user.
4c~eratin~ Environment of the Trainin~~ Exercises
In this illustrative embodiment, each of the training exercises described
above and
more completely below are all or part of one or more computer processes
executing within
a computer system 1000 as shown in Figure 81. Computer system 1000 (Figure 81)
includes a processor 1002 and memory 1004 which is coupled to processor 1002
through
an interconnect 1006. Interconnect 1006 can be generally any interconnect
mechanism for
computer system components and can be, e.g., a bus, a crossbar, a mesh, a
torus, or a
hypercube. Processor 1002 fetches from memory 1004 computer instructions and
executes
the fetched computer instructions. In addition, processor 1002 can fetch
computer
instructions through a computer network 1070 through network access circuitry
1060 such
as a modem or ethernet network access circuitry. Processor 1002 also reads
data from and
writes data to memory 1004 and sends data and control signals through
interconnect 1006
to one or more computer display devices 1020 and receives data and control
signals
through interconnect 1006 from one or more computer user input devices 1030 in
accordance with fetched and executed computer instructions.
Memory 1004 can include any type of computer memory and can include, without
limitation, randomly accessible memory (RAM), read-only memory (ROM), and
storage
CA 02318107 2000-07-20
WO 99/38137 - 14 - PCTNS99/01415
devices which include storage media such as magnetic and/or optical disks.
Memory 1004
includes a trainer 1100 which includes the training exercises described above
and which is
all or part of a computer process which in turn executes within processor 1002
from
memory 1004. Alternatively, trainer 1100 can be implemented as a collection of
computer
processes. A computer process is generally a collection of computer
instructions and data
which collectively define a task performed by computer system 1000.
Each of computer display devices 1020 can be any type of computer display
device
including without limitation a printer, a cathode ray tube (CRT), a light-
emitting diode
(LED) display, or a liquid crystal display (LCD). Computer display devices
1020 each
receive from processor 1002 control signals and data and, in response to such
control
signals, display the received data. In addition, loudspeaker 1020B plays
audible sounds
received from audio processing circuitry 1080. Loudspeaker 1020B can be one or
more
loudspeakers, amplified or unamplified. For example, loudspeaker 1020B can be
a pair of
headphones. Audio processing circuitry 1080 receives digital audio signals and
control
signals from processor 1002 and produces analog audio signals in accordance
therewith for
reproduction through loudspeaker 1020B. Computer display devices 1020, and the
control
thereof by processor 1002, are conventional.
Each of user input devices 1030 can be any type of user input device
including,
without limitation, a keyboard, a numeric keypad, or a pointing device such as
an electronic
mouse, trackball, lightpen, touch-sensitive pad, digitizing tablet, thumb
wheels, joystick, or
voice recognition device. Each of user input devices 1030 generates signais in
response to
physical manipulation by a user and transmits those signals through
interconnect 1006 to
processor 1002.
As described above, trainer 1100 executes within processor 1002 from memory
1004. Specifically, trainer 1100 is all or part of one or more computer
processes executing
within computer system 1000, i.e., processor 1002 fetches computer
instructions of trainer
1100 from memory 1004 and executes those computer instructions. As shown in
Figure
81, trainer 1100 includes the following training exercises: Old MacDonald's
Flying Farm
1102, Circus Sequence 1104, Block Commander 1106, Phonic Match 1108, Phonic
Words
11 I0, Phoneme Identification 1 I 12, and Language Comprehension Builder 1114.
Each of
training exercises 1102-1114 has the respective behavior described above.
Processor 1002,
CA 02318107 2000-07-20
WO 99/38137 PCT/US99/01415
-15-
in executing trainer 1100, (i) presents stimuli such as the stimuli described
above to the
user in the form of video images in the screen views of Figures 1-80 on
display screen 122
and audible sounds through audio processing circuitry 1080 and loudspeaker
1020B, and
('li) receives signals from user input devices 1030 representing the user's
response to such
stimuli. Furthermore, execution of trainer 1100 provides to the user
motivation to continue
use of trainer 1100 in the manner described more completely below.
In this illustrative embodiment, processor 1002 is any of the Pentium series
of
microprocessors available from Intel Corporation of Santa Clara, California
and computer
system 1000 is compatible with the PC personal computer available from
International
Business Machines of Atlanta, Georgia operating under the WINDOWS 95 operating
system available from Microsoft Corporation of Redmond, Washington.
Alternatively,
computer system 1000 can be a Macintosh computer available from Apple Computer
Inc.
of Cupertino, California.
As described briefly above, the user identifies objects represented in display
screen
1022 by clicking on such object. Typically, a user clicks on an object, e.g.,
flying object
102 (Figure 1) by positioning a cursor, e.g., cursor 104, over the object (as
shown in
Figure 2) and actuating a physical button or similar actuation mechanism on a
user input
device such as any of user input devices 1030 (Figure 81). The user moves the
cursor by
physical manipulation one of user input devices 1030. For example, movement of
an
electronic mouse over a generally flat surface is generally detected by the
electronic mouse
either mechanically or optically and the electronic mouse generates signals
representing the
direction and distance of movement of the electronic mouse through
interconnect 1006 to
processor 1002.
Processor 1002 effects corresponding movements of the cursor in display screen
1022 to provide substantially immediate feedback to the user regarding the
current position
within display screen 1022 indicated by the current state of the electronic
mouse. With the
cursor over the object, the user can click on the object by actuating a
physical button on the
electronic mouse to thereby generate and transmit to processor 1002 signals so
indicating.
Clicking on an object by a user is detected by a computer process such as
trainer 1100
(Figure 81) and the computer process takes action in response to the detected
actuation in
accordance with the particular computer instructions which define the behavior
of the
CA 02318107 2000-07-20
WO 99/38137 - 16 - PCT/US99/01415
computer process.
Motivation Mechanisms
In general, a complete training process for a particular user can involve
playing the
above-described training exercises for extended periods of time over a
relatively long
period of time, e.g., approximately one-to-two hours per day, five (S) days
per week, for a
period of 6-8 weeks. In addition, users of the training exercises described
above are
typically children. Children frequently have relatively little appreciation
for the skills and
cognitive abilities developed by the above-described training exercises and,
accordingly,
attention and persistence in adhering to the rigorous schedule described above
is a
significant problem. In addition, the nature of the training exercises as
described briefly
above are not as inherently exciting and motivating as threat/defense games.
Instead of
defending oneself from a simulated threat or racing or fighting an opponent,
the user is
asked to recognize sounds and follow simple directions. For many children,
such
challenges might seem mundane. Accordingly, a number of motivation mechanisms
are
implemented in the training exercises described above to promote use of the
training
exercises by users and to help users maintain and follow the ambitious
schedules
established for proper training. These motivation mechanisms include (i) a
standardized
token economy, (ii) achievement-based context changes, (iii) progress
indicators, (iv)
reward animations using persistent characters and plots, and {v) variable
ratio
reinforcement mechanisms. The motivations mechanisms described herein can be
used
individually or in combination to motivate users to apply themselves and to
maintain the
interest of users throughout the training process.
Standardized Token Economy
In playing the training exercises described above, the user accumulates points
which
can be subsequently exchanged for physical rewards. Such rewards can include
small toys,
T-shirts, recorded music (e.g., tapes and compact discs), and tickets to
sporting events.
The points therefore represent a currency in a token economy. A point counter
such as
point counter 106 (Figure 1) indicates to the user the number of points
accumulated during
any individual play session of a particular training exercise. The represented
accumulation
CA 02318107 2000-07-20
WO 99/38137 - 1,~ - PCTNS99/01415
of points provides motivation of the user to continue to use the training
exercise and to
accumulate points toward a particular reward the user would like.
Such a token economy can provide motivation to play one of the training
exercises
described above rather than others of the training exercises. For example,
Circus Sequence
1104 (Figure 81) and Phonic Match 1108 play relatively quickly, i.e., present
stimuli to and
receive responses from the user in a relatively short time period. The sounds
in question
are very short compared to sounds used in other training exercises. For
example, Circus
Sequence 1104 presents two relatively short frequency sweeps to the user, and
Phonic
Match 1108 presents only single, short phonemes for each response by the user.
By
comparison, LCB 1114 can present audible stimuli in the form of lengthy
sentences such as
"show me the boy is hugging the clown that is kissing the girl" and such
stimuli can be
slowed and stretched and modified to be more easily understood by a user with
LLI.
Attributing the same number of points to each correct response regardless of
the particular
training exercise being played allows players to spend an inordinate amount of
time playing
training exercises with quick stimuli/response exchanges, e.g., Circus
Sequence 1104 and
Phonic Match 1108, and to spend much less time playing training exercises with
slower
stimuli/response exchanges, e.g., LCB 1114. Doing so enables the user to more
quickly
acaunulate poi~s.
Therefore, the number of points accumulated during play of each of the
training
exercises is standardized such that playing different training exercises for
the same amount
of time with the same degree of proficiency accumulates approximately the same
number of
points. As a result, preferences for one training exercise over another based
on point-
earning capacity are removed.
Training exercise 1200 (Figure 82) is a generic representation of any of
training
exercises 1102-1112 of trainer 1100 (Figure 81). Training exercise 1200
includes a number
of field 1202-1212 which define components of the behavior of training
exercise 1200.
Training exercise 1200 further includes animation logic 1214 which presents
reward
animations to the user to periodically reward the user for good performance
during play of
training exercise 1200. Training exercise logic 1216 defines the overall
behavior of training
exercise 1200 described above. For example, if training exercise 1200
represents Old
MacDonald's Flying Farm I 102 (Figure 81 ), then training exercise logic 1214
(Figure 82)
CA 02318107 2000-07-20
WO 99/38137 - 1 g - PCT/US99/01415
defines the behavior described above with respect to Old MacDonald's Flying
Farm 1102.
In general, each of training exercises 1102-1114 (Figure 81) presents the user
with
a reward animation after a number of correct responses have been provided by
the user. A
number of points are awarded for each correct response to stimuli and a number
of bonus
points are awarded each time the user has earned a reward animation. To
standardize the
token economy, training exercise 1200 (Figure 82) specifies in point increment
field 1202,
animation interval field 1204, and bonus field 1206 the amount of points a
user can earn
during play of the training exercise represented by training exercise 1200.
Specifically, point increment field 1202 specifies the number of points
awarded to
the user for each correct response. The number of points is directly related
to the amount
of time typically required to present the user with stimuli and to receive
corresponding
response signals from the user through user input devices 1030 (Figure 81).
This time is
referred to herein as a transaction time.
As described above, Circus Sequence 1104 and Phonic Match 1108 typically have
the smallest transaction times. Accordingly, the number of points awarded for
each correct
response during play of Circus Sequence 1104 and Phonic Match 1108 as
represented in
point increment field 1202 (Figure 82) is relatively low. In this illustrative
embodiment,
two (2) points are awarded for each correct response in Circus Sequence 1104
and two (2)
points are awarded for each correctly identified pair of panes in Phonic Match
1108.
Circus Sequence 1104 represents the user's points accumulated during the
current session
in a point counter box 310 (Figure 17). Similarly, Phonic Match 1108 (Figure
81)
represents the user's points accumulated during the current session in a point
counter box
506 (Figure 41).
Phoneme Identification 1112 (Figure 81 ) has a typical transaction time which
is
slightly greater than the typical transaction times of Circus Sequence 1104
and phonic
Match 1108. Therefore, if training exercise 1200 (Figure 82) represents
Phoneme
Identification 1112, point increment field 1202 stores data indicating that
three (3) points
are awarded to the user for each correct response. Phoneme Identification 1112
represents
the user's points accumulated during the current session in a point counter
box 706 (Figure
59).
Old MacDonald's Flying Farm 1102 has a typical transaction time which is
greater
CA 02318107 2000-07-20
WO 99/38137 _ 19 _ PCT/US99/01415
than that of Phoneme Identification 1112. Accordingly, point increment field
1202 stores
data indicating that four (4) points are awarded to the user for each correct
response if
training exercise 1200 represents Old MacDonald's Flying Farm 1102. Old
MacDonald's
Flying Farm 1102 represents the user's points accumulated during the current
session in a
point counter box 106 (Figure 1).
Phonic Words 1110 (Figure 81) has a typical transaction time which is greater
than
the typical transaction time of Old MacDonald's Flying Farm 1102 and less time
than the
typical transaction time of LCB 1114. Five (5) points are therefore awarded to
the user for
each correct response if training exercise 1200 (Figure 82) represents Phonic
Words 1110.
Phonic Words 1110 represents the user's points accumulated during the current
session in a
point courner box 606 (Figure 49).
LCB 1114 (Figure 81) and Block Commander 1106 generally have the greatest
transaction times. Therefore, if training exercise 1200 (Figure 82) represents
LCB 1114 or
Block Commander 1106, point increment field 1202 stores data indicating that
six (6)
points are awarded to the user for each correct response. LCB 1114 represents
the user's
points accumulated during the current session in a point counter box 906
(Figure 70).
Similarly, Block Commander 1106 (Figure 81 ) represents the user's points
accumulated
during the current session in a pointer counter box 406 (Figure 35).
Thus, the user accumulates three (3) times as many points for each correct
response
in LCB 1114 (Figure 81) as are accumulated for each correct response of Circus
Sequence
1104. Such is appropriate since LCB 1114 requires approximately three (3)
times as much
time to present stimuli to the user and receive the user's response as does
Circus Sequence
1104. Therefore, the number of points accumulated during play of each training
exercise is
normalized over time. In other words, playing each training exercise for a
period of time
with substantially equivalent levels of proficiency accumulates approximately
equal
amounts of points.
As described briefly above and more completely below, the user is periodically
rewarded with an animation. Animation interval field 1204 (Figure 82) stores
data
representing the number of correct responses required to earn a reward
animation. The
number of correct responses is referred to herein as the animation interval
and, like the
point increment, can be related to the transaction time. In the case of the
animation
CA 02318107 2000-07-20
WO 99/38137
PCT/US99/01415
-20-
interval, the relation is generally an inverse re:ation in that a larger
animation interval
corresponds to a smaller transaction time and vice versa.
When the user earns a reward animation through correctly responding to stimuli
the
number of times represented in animation interval field 1204, the user is also
awarded a
number of bonus points. The number of bonus points awarded at each animation
interval is
represented by data stored in bonus field 1206. The bonus points generally
serve two
purposes. First, the bonus points provide a relatively short term, achievable,
and desirable
goal for the user. Second, the bonus points can be adjusted to fiuther
equalize point
accumulation among the respective training exercises over time. Various
combinations of
animation intervals and bonus points can effect accumulation at non-integer
rates per
correct response by the user to give added flexibility in equalizing point
accumulation
among training exercises 1102-1114 while maintaining the simplicity of integer
point values
and relatively small point increments which are relatively easily grasped and
appreciated by
users. For example, an animation interval of twenty (20) correct responses and
a bonus of
ten (10) points per reward animation in effect results in one-half point per
correct response
by the user over time.
Thus, point increments, animation intervals, and bonus points as represented
in
point increment field 1202, animation interval field 1204, and bonus field
1206,
respectively,. can be used to equalize point accumulation over time among the
various
training exercises. The following table sets out the point increments,
animation intervals,
and bonuses of the various training exercises of the illustrative embodiment.
1 A
Training exercise Point Animation Bonus Points Effective
Increment Interval per Reward Point
Animation Incr't
Old MacDonald's 4 10 10 50
Flying Farm
Block Commander 6 S 0 30 6
Circus Sequence 2 10 10 30 3
Phonic Match 2 ~ varvinn ~ ~~rW~n
CA 02318107 2000-07-20
WO 99/38137 - 21 - PCT/US99/01415
Training ezercisePoint Animation Bonus Points Effective
Increment Interval per Reward
Point
Animation
Incr't
Phonic Words 5 10 10
Phoneme 3 10 10 40 4
Identification
a 6 9 10 64 7.11111
Comprehension
Builder
The point increments, animation intervals, and bonuses shown in Table A
provide a
relatively standardized token economy such that users generally do not prefer
one training
exercise over another. Statistics regarding the amount of time users spend
playing each of
the training exercises can be collected in the manner described in the
Compliance
Monitoring Application, which description is incorporated herein by reference,
and
disparity in the amount of time spent by users using each training exercise
call be detected.
Upon such detection, the values shown in Table A can be adjusted so as to
remove any
~sParitY in the tune spent using the respective training exercises by users.
Prom~ess ~~c.~tors
As described briefly above, the user is rewarded periodically with a reward
animation. Each of the training exercises described above provides a progress
indicator to
convey to the user information regarding the progress toward such a reward
animation.
Such progress indicators are best described by way of example.
Old MacDonald's Flying Farm 1102 includes a progress indicator 108 (Figure 3)
which is in the form of a grandmother driving a tractor across a farm scene.
For each
correct response given by the user in response to a detected change in the
repeated
phoneme, progress indicator 108 moves a predetermined offset increment from
right to left.
Figure 5 shows progress indicator 108 moved further to the left as a result of
one or more
additional correct responses. Progress indicator 108 is an integral component
of the scene
in which Old MacDonald's Flying Farm takes place. Specifically, the
grandmother-driven
tractor is an integral component of the farm depicted in Figures 1-lb.
CA 02318107 2000-07-20
WO 99/38137 - 22 - PCTNS99/01415
In addition, progress indicator 108 is animated. Specifically, a brief
progress
animation gives both the grandmother and the tractor motion in moving from one
position
to the next. The motion is relatively smooth and provides an appearance of
progress. By
contrast, simply removing a progress indicator from a previous position and re-
drawing the
progress indicator at a next position gives an appearance of an instantaneous
change of
state rather than a smooth progression. The progress animation is entertaining
and
pleasing, providing a continuous rate reinforcement mechanism. In other words,
each
correct response is met with immed;ate feedback in the form of a brief
animation moving
progress indicator 108 and the feedback is desirable to the user, motivating
the user to earn
another small progress animation. In addition, correct responses elicit an
immediate,
pleasant sound such as a "ding." Conversely, an incorrect response elicits a
less pleasing
sound, such as a "thank," to indicate to the user that a response was received
but the
response was not the response sought by the training exercise.
When the user has earned a reward animation by responding correctly a number
of
times represented in animation interval field 1204 (Figure 82), progress
indicator 108
(Figure 5) has moved to its left-most position and the user has achieved a
short-term goal.
Accordingly, a reward animation is presented to the user. For example, a cloud
116
(Figure 6) darkens, a weathervane 112 pulls out an umbrella 114 and raindrops
fall from
cloud 116. The rain then stops and flying object 102 (Figure 7), which is
represented as a
cow with a rocket pack, flies by and takes umbrella 114 off the scene.
Circus Sequence 1104 (Figure 81) includes a progress indicator 312 (Figure 17)
which is raised from the bottom of the circus scene to a top position by an
increment with
each correct response by the user. Progress indicator 312 is represented as a
wheeled
circus cage raised by a rope and is animated in that eyes move back and forth
as the cage is
raised from one position to the next. Figure 19 shows progress indicator 312
in a slightly
raised position resulting from one or more correct responses by the user. When
the user
responds correctly a number of times represented in animation imerval field
1204 (Figure
82), the user is rewarded with a reward animation as shown in Figures 20-22.
This
P~~~' animation is described more completely below.
After each animation, the progress indicator of Circus Sequence 1104 changes
to
provide variety to the user. Other progress indicators include a woman on a
trapeze, a man
CA 02318107 2000-07-20
WO 99/38137 - 23 - PCT/US99/01415
climbing a rope, a hippopotamus raised by a hot-air balloon (progress
indicator 312B in
Figure 30), and a merry-go-round (progess indicator 312C in Figure 32).
Phonic Match 1108 includes a progress indicator 508 (Figure 41 ) which is
represented as a roadrunner skating with in-line skates and wearing a top hat.
Progress
indicator 508 is perpetually animated in that progress indicator 508 appears
to be skating
without pausing between user selection of individual ones of panes 504. Each
time a
matching pair of panes 504 is correctly identified by the user, progress
indicator 508 moves
one position lower toward a sidewalk on which progress indicator 508
ultimately skates.
Phonic Match 1108 operates slightly different from the other training
exercises
described herein in that the animation interval is variable. Phonic Match 1108
determines a
maximum number of tries the user should need to correctly identify all
matching pairs of
panes 504 and awards the user with bonus points for using fewer tries to
correctly identify
all matching pairs. The processing of Phonic Match is illustrated as logic
flow diagram
1300 (Figure 83).
In step 1302, Phonic Match 1108 (Figure 81) determines the expected maximum
number of tries. The expected maximum number of tries depends on the number of
panes
504 (Figure 41 ) and an estimation of the memory capacity of the user. In one
embodiment,
Phonic Match 1108 (Figure 81) empirically determines the mean and standard
deviation of
the number of tries required to correctly identify all matching pairs of panes
504 (Figure
41) and sets the expected maximum number of tries as the mean number of tries
plus one
standard deviation. Phonic Match 1108 makes such an empirical deternlination
by
assuming the user can remember the phonemes of two {2) of panes 504 at any one
time and
randomly trying individual ones of panes 504, using the simulated memory, and
counting
the number of tries to correctly identify all matching pairs of panes 504.
Through repeated
simulations of this type, phone Match 1108 can empirically determine an
estimated mean
and standard deviation. While, in this illustrative embodiment, simulated
trying assumes a
memory of two (2) of panes 504, other assumptions can be made with respect to
the user's
memory capacity. For example, assuming a memory of one or even zero can be
used to
train memory impaired individuals.
In step 1304 (Figure 83), Phonic Match 1108 (Figure 81) repeatedly receives a
try
in the form of signals from the user identifying a selected one of panes 504
(Figure 41 ). In
CA 02318107 2000-07-20
WO 99/38137 - 24 - PCTNS99/01415
test step 1306 (Figure 83), Phonic Match determines with the try correctly
idemifies a
matching pair of panes 504 (Figure 41 ), i. e., whether the identified pane
matches the most
recently identified pane. If not, Phonic Match 1108 (Figure 81) waits for the
next try. If
so, Phonic match 1108 removes the matching pair from the display in step 1308
(Figure 83)
and moves progress indicator 508 (Figure 41) down one position in step 1310
(Figure 83).
The total number of positions of progress indicator 508 (Figure 41) is equal
to the expected
maximum number of tries determined in step 1302 (Figure 83). Figure 42 shows
that a
matching pair of panes 504 have been removed and progress indicator 508 has
been
lowered one or more positions. After step 1310 (Figure 83), Phonic Match 1108
(Figure
81 ) waits for the next try by the user.
When all matching pairs have been correctly identified by the user, processing
transfers to step 1314 (Figure 83) in which steps 1316 and 1318 are performed
repeatedly
until position indicator 508 (Figure 42) is positioned at the bottom. In step
1316 (Figure
83), the user is awarded a number of bonus points, and, in step 1318, position
indicator 508
(Figure 42) is lowered one position. Therefore, the amount of bonus points
awarded to the
user is proportional to the number of tries less than the expected maximum the
user
required to correctly identify all matching pairs of panes 504. If position
indicator 508 is
already at the bottom. before step 1314 (Figure 83), i.e., before the user
correctly identifies
all matching pairs of panes 504 as shown in Figure 44, no bonus points are
awarded to the
user. Accordingly, the user has a strong incentive to correctly identify all
matching pairs of
panes 504 in as few tries as possible.
Block Commander 1106 (Figure 81 ) uses non-animated discrete progress
indicators
408 (Figure 36) to indicate to the user progress toward a reward animation.
Progress
indicators 408 (Figure 37) are different yet belong to a theme. In the
illustrative example
of Figure 37, the theme is one of milk. Since each of progress indicators 408
can be
different, the user is motivated to see the various forms of representation of
progress
indicators 408 belong to the theme of milk. Specifically, Block Commander 1106
(Figure
81 ) randomly selects a graphical image from a collection of graphical images
of a given
theme each time the user responds correctly to stimuli presented by Block
Commander
1106. When the user has provided a number of correct responses equal to the
animation
interval represented in animation interval 1204 (Figure 82), when training
exercise 1200
CA 02318107 2000-07-20
WO 99/38137 - 25 - PCT/US99/01415
represents Block Commander 1106, animation logic 1214 presents the user with a
reward
animation as shown in Figure 38. Specifically, balloons 410 are shown to float
up from
blocks 404 while a cat 412 bats a paw at balloons 410.
After presentation of a reward animation, Block Commander 1106 changes the
theme of progress indicators. For example, Figure 39 shows progress indicators
408B in
the form of mouse-theme jack-in-the-boxes. The associated reward animation
includes fish
410B (Figure 40) floating up the scene and blowing bubbles.
Phonic Words 1110 (Figure 81) uses discrete progress indicators 610 (Figure
48) to
indicate to user progress toward a reward animation. Each of progress
indicators 610
changes its state, e.g., from a representation of an egg to a bird 610A
(Figure 49), in
response to a correct response by the user. The change of state is animated to
provide an
appearance of smooth progress. For example, in transition from an egg
representation to
bird representation 610A, progress indicators 610 are animated to show the egg
cracking
and a bird emerging as shown as progress indicator 610B (Figure 50). When the
user has
provided ten ( 10) correct responses, all progress indicators 610 (Figure 5 i
) are shown as
bird representations and the user is rewarded with a reward animation.
Specifically, egg
shells 614 (Figure 52) float down the scene and enclose progress indicators
610 an slink
away as a single inch worm as shown in Figure 53.
To avoid monotony and to maintain the user's interest in Phonic Words 1110
(Figure 81 ), progress indicators 610 are replaced with progress indicators 61
OC (Figure
54) after the reward animation. Progress indicators 610C are shown as honey
jars and the
progress animation includes a honey bee 616 which flies from one honey jar
into the next to
eat the honey as shown in Figure 55. The following other progress indicators
provide
variety and some surprise and enjoyment to the user when playing Phonic Words
1110.
Progress indicators 610D (Figure 56) spring up as flowers as a squirrel 618
moves from
one of progress indicators 610D to the next. Progress indicators 610E (Figure
57) are
represented as lily pads and the progress animation includes a frog 620 which
jumps from
one of progress indicators 610E to the next. Progress indicators 610F (Figure
58) are
represented as carrots or holes in the ground and the progress animation
includes a rabbit
622 which pulls a carrot from the ground and sits in the resulting hole when
progressing
from one of progress indicators 610F to the next.
CA 02318107 2000-07-20
WO 99/38137 - 26 - PCT/US99/01415
Phoneme Identification 1112 (Figure 81 ) uses discrete progress indicators 708
(Figure 60) to indicate to the user progress toward a reward animation. Each
of progress
indicators 708 changes its state, e.g., from a representation of an unlit
torch to a lit torch
with a flame 712 (Figure 61), in response to a correct response by the user.
The change of
state is animated to provide an appearance of smooth progress. For example,
the torch
lights smoothly, progressing from a small flame to a full size flame 712 and
continues to
flicker throughout use of Phoneme Identification 1112.
LCB 1114 (Figure 81 ) uses discrete progess indicators 910 (Figure 70) to
indicate
to the user progress toward a reward animation. Each of progress indicators
910 changes
its state, e.g., from a blank representation to a representation of an item
916 (Figure 71), in
response to a correct response by the user. Item 916 comports with a theme,
e.g., items
found in an office. Progress indicators 910 (Figure 72) include a computer, a
desk, a fan, a
file cabinet, a potted plant, a chair, a stapler (Figure 73), a phone, and a
waste-paper
basket. Progress indicators 910 are component objects of an associated reward
animation.
Specifically, the associated reward animation includes the objects of progress
indicators
910 floating in space (Figure 74) and coming together (Figure 75) to form a
make-shift
space vehicle (Figure 76) which is flown out of the scene. The objects of
progress
indicators 910 are not ones ordinarily associated with space vehicles.
Accordingly,
elements of surprise and curiosity motivate the user to provide correct
responses to stimuli
to thereby earn objects as progress indicators 910 to see what type of space
vehicle can be
created therefrom.
While progress indicaxors 910 include objects of a theme, it is possible that
objects
which are not so easily categorized according to a theme can be combined is a
curious and
surprising manner to form a space vehicle reward animation. For example,
progress
indicators 9108 (Figure 77) include sports objects such as a baseball, a cap,
and a baseball
glove and non-sports objects such as a ballpoint pen and a steam iron.
Progress indicators
910B even include objects not found in reality and not easily recognized,
These objects
come together (Figure 78) to form a space vehicle (Figure 79) which is flown
ogthe scene
(Figure 80).
Variable Ratio Reinforcement
CA 02318107 2000-07-20
WO 99!38137 - 27 - PCT/US99/01415
The progress indicators described abcve are a fixed ratio reinforcement
mechanism
in that the reinforcement is provided in response to a fixed number of correct
responses.
Such mechanisms sometimes provide limited motivation. For example, if the user
wishes to
see flying object 102 (Figure 4) fly into home 110, the user has only to
provide one correct
response to stimuli presented by Old MacDonald's Flying Farm. The user can
provide the
next correct response whenever the user wishes to see flying object 102 fly
into home 110
again. Reward animations described above and more completely below provide a
more
entertaining reward but delay the reward until a number of correct responses
are given by
the user. Therefore, if the user wishes to see the space vehicle of Figures 74-
76 come
together and fly offthe scene, the user must respond correctly to stimuli nine
(9) times.
Therefore, the user is motivated to provide nine (9) correct responses but
loses that
motivation once the reward animation is presented until the user wishes to see
another
space vehicle come together and fly off the scene.
Variable ratio reinforcement provides rewards to the user at random intervals.
The
random, unpredictable nature of variable ratio reinforcement provides a degree
of surprise
and excitement which motivates the user. Since the user doesn't know when the
reward is
coming, the user is motivated to provide as many correct responses as possible
to thereby
increases the likelihood that variable ratio reinforcement is presented to the
user.
Phoneme Identification 1112 (Figure 81) provides a good example of variable
ratio
reinforcement. As described above, each correct response by the user causes a
torch of
progress indicators 708 (Figure 61 ) to light. In addition, at random
intervals, one of
progress indicators 708 is animated in a different manner. For example, a
correct response
by the user can cause an ice cream scoop 714 (Figure 62) to spring up on a
torch of
progress indicators 708. Similarly, a cup of hot chocolate 716 (Figure 64) can
spring up on
a torch of progress indicators 708 in response to a correct response by the
user. Such
unexpected rewards are sometimes referred to as nuggets. Other nuggets of
Phoneme
Identification 11 I2 include many other small animations including bouquets of
flowers and
apples. The diversity of such small animations adds to the surprise and
entertainment and
piques the user's curiosity to thereby motivate the user to provide additional
correct
responses to stimuli. Variable ratio reinforcement is, in this illustrative
embodiment,
randomly distributed with a uniform distribution and a mean appearance ratio
of about 3%.
CA 02318107 2000-07-20
WO 99/3813? PCT/US99/01415
-28-
In other words, nuggets appear approximatel;~ 3% of the time. Significantly
greater
appearance ratios than 3%, e.g., greater than 10%, don't motivate users as
much and the
effect becomes closer to that of reward animations in that the user knows a
nugget is
coming soon. Significantly lesser appearance ratios than 3%, e.g., less than
1%, don't
motivate users as much since the appearance of a nugget is so rare that users
don't expect a
nugget any time soon enough to be motivated to reach the nugget.
Old MacDonald's Flying Farm 1102 (Figure 81 ) also uses variable ratio
reinforcement to motivate users. Typically, flying object 102B (Figure 9)
flies into home
110 in response to a correct response by the user in the manner shown, i.e.,
with a rocket
pack on the back of the cow. At random intervals, flying object 102B enters
home 110 in
different ways. For example, the cow jumps onto the rocket packet and "surfs"
into home
110 as shown in Figure 10. Alternatively, the cow of flying object 102B rolls
over and
reclines on the rocket pack to ride into home 110 as shown in Figure 11 or
lands outside
home 110 and runs along the ground into home 110 with an accompanying
"clumpity-
clumpity" sound of hooves running on the ground.
Similarly, at higher levels of Old MacDonald's Flying Farm 1102 which are
described fiuther below, the normal feedback animation of a correct response
is flying of a
flying object 202 (Figure 13) into a home 210 (Figure 15) in the form of a
chick suspended
from a hot-air balloon flying under a mother hen. At random intervals, flying
object 202
enters home 210 in different ways. For example, other chicks 212 can be
already present
under home 210, i. e., the mother hen.
LCB 1114 (Figure 81) also uses variable ratio reinforcement to further
motivate
users. A young girl astronaut 912 (Figure 71 ) sits still while the user
responds to stimuli
preserved by LCB 1114 and becomes the pilot of the space vehicle ultimately
assembled
from the objects earned by the user from correct responses as shown in Figures
75-76. At
random intervals, astronaut 912 rises up in apparent zero-gravity as shown in
Figure 72 and
spins and returns to her original position as shown in Figure 73. To see
astronaut 912 rise
and spin again, the user has to respond to stimuli correctly some unknown,
random number
of times and the uncertainty of the number of times required to see astronaut
912 rise and
spin provides motivation to the user to respond correctly and to do so
quickly.
CA 02318107 2000-07-20
WO 99/38137 PCT/US99/01415
-29-
As described above, each training exercise presents the user with a reward
animation periodically. Each reward animation is designed to be entertaining
such that the
user is motivated to continue to respond correctly to stimuli to see the next
reward
animation. Entertainment value of reward animations comes from humor and
surprise
incorporated into the reward animations. The user's interest in reward
animations is
encouraged even further by inclusion in the reward animations of persistent
characters and
plots. The following examples are illustrative.
Figures 20-22 show individual screen views of a reward animation of Circus
Sequence 1104 (Figure 81). A character 320 (Figure 20) in the form of a dog
magician is
about to perform the infamous sawing in half of another character 322, e.g., a
monkey,
lying in a box 324. In Figure 21, character 320 is sawing character 322 in
half through box
324, and character 320 separates the sawed halves of box 324 (and of character
322) in
Figure 22. This reward animation is entertaining in and of itself. However,
characters 320
and 322, and the plot, persist and continue through a subsequent reward
animation shown
in Figures 23-26.
The user continues to use Circus Sequence 1104 until the user earns another
reward
animation. The reward animation of Figures 23-26 can immediately follow the
reward
animation of Figures 20-22 as the next reward animation earned by the user.
Alternatively,
one or more other reward animations can be earned between the reward
animations of
Figures 20-22 and 23-26. However, since the reward animation of Figures 23-26
continues
the persistent plot involving consistent characters of the reward animation of
Figures 20-
22, animation logic 1214 (Figure 82) of Circus Sequence 1104 (Figure 81)
ensures that the
reward animation of Figures 20-22 is presented to the user prior to
presentation of the
reward animation of Figures 23-26.
When another reward animation is earned by the user, character 320 (Figure 23)
waves his hands and wand and causes the feet 326 of character 322 to change to
the feet of
a pig. Character 320 continues to wave his hands and wand while feet 326
change fiuther
to feet 326B of a sheep (Figure 24), feet 326C of an elephant (Figure 25), and
feet 326D of
a woman (Figure 26) at which point the sound of someone in the circus audience
whistling
is played and character 322 blushes. After the reward animation of Figures 23-
26, the user
CA 02318107 2000-07-20
WO 99/38137 - 3~ - PC'T/US99/01415
sees that characters and plots in the reward ar_~.mations persist and
continue. As a result,
the user becomes curious as to what happens next in the continuing plot and
what happens
next to the persistent, consistent character. The user is therefore motivated
to earn another
reward animation.
Upon subsequent earning of a reward animation by the user through correct
responses to stimuli, the reward animation of Figures 27-29 continues the plot
and
characters of the reward animations of Figures 20-26. Specifically, character
320 (Figure
27) reads a book on magic while a new character 328 enters the scene.
Character 328,
e.g., a clown, pushes the sawed halves of box 324 and character 322 together
(Figure 28)
while character 320 continues to read. Character 322, now whole again, stands
and walks
off scene, still with a woman's feet, while character 320 reads and is
oblivious to the
happenings behind.
Thus, reward animations of Figures 20-22, 23-26, and 27-29 collectively form a
single composite reward animation whose individual component animations are
presented
in series. The component reward animations of a composite reward animation can
be
presented contiguously, i.e., in immediately successive reward animation
intervals, as
earned by the user by correctly responding to stimuli, without interceding
reward
animations. Alternatively, component reward animations of a composite reward
animation
can be presented in reward animation intervals which are separated by one or
more
interceding reward animation intervals in which unrelated reward animations
are presented
to the user. For example, an unrelated reward animation can be presented to
the user after
the reward animation of Figures 23-26 and before the reward animation of
Figures 27-29.
However, it is preferred that component reward animations of a composite
reward
animation are presented is sufficient temporal proximity that the user can
recollect the
characters and plot of previously presented component reward animations such
that the
context of a currently presented component reward animation is fully
appreciated.
Whether contiguous or not contiguous, it is preferred that component reward
animations of a composite reward animation are presented in sequence and that
all earlier
component reward animations are presented to the user as reward animations are
earned
prior to presentation of a subsequent component reward animation. For example,
it is
preferred that the reward animations of Figures 20-22 and 23-26 are presented
to the user,
CA 02318107 2000-07-20
WO 99/38137 - 31 - PCT/US99/01415
with or without interceding unrelated reward animations, prior to presentation
to the user
of the reward animation of Figures 27-29 so that the user can fully appreciate
the context
of the latter reward animation.
Characters of reward animations can persist beyond a single plot. For example,
in a
reward animation having a distinct and unrelated plot to the reward animations
of Figures
20-29, character 320 (Figure 33) has trained another character 332, e.g., a
leopard, to jump
through a hoop 334 dipped in bubble solution 336. After jumping through hoop
334,
character 332 is encapsulated in a bubble and flies about the circus (Figure
34). Upon
recognizing character 320, the user is reminded of the prior entertaining
composite reward
animation and anticipates subsequent reward animations continuing the plot of
the reward
animation of Figures 33-34 and subsequent reward animations which feature the
persistent
personality and appearance of character 320 with which the user is becoming
familiar.
Such anticipation further motivates the user to earn subsequent reward
animations by
responding promptly and correctly to stimuli presented by Circus Sequence
1104.
~Pment-Based Context Chances
The motivation mechanisms described above provide short term and medium term
motivation for the user to respond quickly and promptly to stimuli presented
by various
training exercises. Longer range motivation is provided and freshness is added
to the
user's experience by periodically changing context based on achievement of the
user. As
described above, the training program of the user involves extended use of the
training
exercises described above over a period of weeks. The essence of each training
exercise
remains the same during the entire period such that language comprehension
skills of the
user are fully developed rather than merely tested. However, the visual
context of various
training exercises can change in accordance with the user's achievement
without departing
from the essence of such training exercises.
In this illustrative embodiment, each training exercise has a number of
categories, a
number of durations for each category, and a number of levels for each
duration of each
category. Each level includes parameters which control a degree of difficulty
in the user's
use of a particular training exercise. For example, each level can specify a
duration of an
inter-stimuli interval (ISI) which is shorter for higher degrees of
difficulty. In other words,
CA 02318107 2000-07-20
WO 99/38137 - 32 - PCT/US99/01415
as the user performs better in responding to s~imuli, the period between two
stimuli is
shortened to further challenge the user's ability to properly perceive the
stimuli.
The adjustment of levels is described more completely in the Training
Application.
Briefly, the user's level for a particular one of training exercises 1102-1114
(Figure 81) is
incremented each time the user responds to stimuli correctly a predetermined
number of
consecutive times, e.g., three (3). Each time the user responds incorrectly to
stimuli, the
user's level for the particular training exercise is reduced. Accordingly,
each training
exercise quickly adapts to the abilities of the user and challenges the user
within the user's
abilities without overwhelming the user with challenges beyond the user's
ability. The level
at which a particular training exercise challenges the user rises relatively
quickly to the
user's ability and oscillates about the level at which is the user is capable
of responding
correctly. When the level of the particular training exercise has dropped a
predetermined
number of times, e.g., eight (8), the training exercise switches to a
different category. One
result of this quickly adapting level mechanism is that users respond
correctly to stimuli
about 80% of the time regardless of the ability of the user. Progress of the
user is
measured in the Ievel achieved rather than the percentage of correct
responses.
Each duration is also associated with parameters which control a degree of
diflxculty in the user's ability to respond correctly to stimuli of a
particular training exercise.
For example, each duration can specify a duration of the stimuli themselves
which shortens
as difficulty increases. Thus, as a user masters accurate perception of a
particular duration
of stimuli, the stimuli are shortened to further challenge the user's
perception thereof.
As described briefly above, each of training exercises 1102-1114 (Figure 81)
switches between categories periodically. Each category specifies a specific
perception of
the user to train. For example, those training exercises which challenge and
improve the
user's ability to distinguish similarly sounding phoneme defines a pair of
similarly sounding
but distinct phonemes for each category. Such pairs include, for example, "ba"
and "da,"
"si" and "sti," and "shu" and "chu." LCB 1114 includes categories which
specify specific
linguistic constructs such as negation, male/female pronouns, and past tense.
The level at which a user uses a particular training exercise changes
relatively
frequently in the manner described above. However, durations and categories
change
significantly less frequently. Therefore, changing of a duration represents a
milestone in the
CA 02318107 2000-07-20
WO 99/38137 - 33 - PCTNS99/01415
achievement of the user. To provide an indication of long-term progress to the
user,
training exercise contexts change with a changing of the duration achieved by
the user.
Figures 1-7 show screen views of Old MacDonald's Flying Farm 1102 at a
particular duration. Figures 8-11 show play of Old MacDonald's Flying Farm
1102 at a
different duration. In Figures 8-11, flying object 102 is replaced with flying
object 102B.
The cow and rocket pack of flying object 102B have different colors to
indicate to the user
that a new level of achievement has been reached. When a new category is
selected, the
context of Old MacDonald's Flying Farm 1102 changes more significantly. As
shown in
Figures 12-16, the barn yard of Figures 1-11 is replace with a chicken coup.
Flying objects
102 and 1028, which are cows flying around by mean of rocket packs on their
backs, are
replaced with flying object 202, which is a chick flying around in half of an
eggshell
suspended from a hot-air balloon. Accordingly, home 110, which is a barn, is
replaced
with home 210, which is a mother hen. In one embodiment, a change in context
is heralded
by a scene change animation. For example, to change from the barn yard scene
of Figure
11 to the chicken coup of Figure 12, the scene of Figure 11 shrinks within a
cloud-
scalloped shaped border to reveal that the barn yard scene is a dream of a
bulldog 120
which briefly wakes and returns to sleep to begin dreaming of the chicken coup
scene of
Figure 12 which expands to fill the entirety of display screen i 022 (Figure
81 ) as shown in
Figure 12. The scene change animation provides rewarding entertainment to the
user and
draws attention to the achievement of the user in progressing to the next
category.
Circus Sequence 1104 (Figure 81) provides achievement-based context in the
form
of a type of animal 304 (Figure 17) at the entrance 306, the color of entrance
306 and the
color of ring 308. Duration is indicated by the color of entrance 306 and ring
308, change
of which is heralded by a scene change animation. For example, in a scene
change
animation shown in Figures 30-32, animals 3048 (Figure 30) are removed (Figure
31)
while a janitor character 330 paints ring 3088 to have a new color. Entrance
3068 is
replaced with entrance 306C (Figure 32) and animals 3048 are replaced to
create the new
context. Category changes are heralded by a similar scene change animation in
which one
type of animal, e.g., animals 304B in the form of giraffes, replace a
different type of animal,
e.g., animal 304 (Figure 17) in the form of a lion.
Phonic Match 1108 (Figure 81) provides achievement-based context in the form
of
CA 02318107 2000-07-20
WO 99/38137 - 34 - PCT/US99/01415
progress indicator 508 (Figure 41), an associated reward animation, and point
counter 506.
At one category, progress indicator 508 is represented as an animated skating
roadrunner
wearing a top hat, the various reward animations involve the roadrunner
skating in one
form or another, and point counter 506 is represented as a lit billboard. The
reward
animations include animations of progress indicator 508 skating and being
drenched by a
car driving through a nearby puddle, being blown backwards and out of the
scene by a fan,
and getting into a car to be driven away rather than skating. At a higher
category, progress
indicator 508 is replaced with progress indicator 508B (Figure 47) which is
represented as
a duck which spins in a progress animation and which incrementally moves down
toward a
beach chair. Reward animations include the beach chair falling backward and
throwing
progress indicator 508B into the distant background, progress indicator 508B
jumping on
the beach chair as if it were a trampoline and bouncing off scene, and
progress indicator
508B swimming a back-stroke along the bottom edge of the scene.
Phoneme Identification 1112 (Figure 81 ) provides achievemem-based context in
the
form of a venue in which the training exercise is played. At one category, the
venue is a
snow field and characters such as characters 710A-B (Figure 60) are shown
facing one
another on snow skis. At a higher category, the venue is an indoor arena,
(Figure 67).
The context changes with changes in duration and category do not change the
manner in which the respective training exercises are played. Instead, the
changing context
provides the user with an indication of progress through the various durations
and
categories of each training exercise.
Im rovement of User Attention
Use of the training exercises described above has been shown to significantly
improve attentiveness of users and the ability of users to focus on a task at
hand.
Specifically, standard measures of such attentiveness and cognitive
flexibility such as the
known Conner's rating scale have shown significant improvement in certain
users. Such
users include those with such disorders as attention deficit disorder (ADD),
attention
deficit disorder with hyperactive disorder (ADD/I~), appositional behavior,
and conduct
disorder. In particular, use of the training exercises above reduce pulsivity,
distractability,
and hyperactivity and simultaneously increase the user's ability to focus and
sustain
CA 02318107 2000-07-20
WO 99/38137 _ 35 _ PCT/US99/01415
attention and to switch cognitive sets.
As described briefly above and more completely in the Training Application,
the
level at which the user is challenged is relatively highly adaptive to the
particular abilities of
the user such that each user is able to respond correctly to about 80~0 of the
stimuli. Such
has particular advantage to training users with attention disorders: Since the
level of
diiixculty adapts to the ability of the particular user, users with lesser
ability are not
frustrated with a significant lack of positive reinforcement. From a different
perspective,
users with unusually low levels of ability are given ample positive
reinforcement with
sufficient frequency to motivate such users to continue with the training
exercises.
Maintenance of interest in training exercises is essential in training users
with attention
disorders.
In addition, the motivation mechanisms described above use no negative
reinforcement. The "thank" audible feedback in response to an incorrect
response by the
user is primary to acknowledge to the user that a response has been received
by the training
exercise. Negative reinforcement tends to fiustrate users with attention
disorders such as
those listed above. Incorrect responses generally result in no more than the
absence of
positive reinforcement in the form of motivation mechanisms described above.
The behavior of the training exercises described above also require activation
of
some sort by the user. Accordingly, nothing happens until the user indicates
the user is
ready to perceive stimuli. For example, Old MacDonald's Flying Farm 1102
(Figures 81
and 1) plays no phonemes for the user until the user grabs flying object 102
(Figure 2).
Umil then, flying object 102 continues to fly about the farm scene. Similarly,
Circus
Sequence 1104 (Figure 81) presents a listening button 302 (Figure 17) to the
user and does
not present any up or down sweeps to the user until the user presses listening
button 302
using conventional user interface techniques. Phonic Words 1110 (Figure 81)
also includes
a listening button 612 (Figure 48) which must generally be pressed by the user
before
Phonic Words 1110 will present any stimuli to the user. Phoneme Identification
11 i2
(Figure 81 ) requires that the user press a similar listening button (Figure
62) prior to
presentation of any stimuli by Phoneme Identification 1112. LCB 1114 (Figure
81)
similarly includes a listening button 914 (Figure 70) which the user must
generally press
before LCB 1114 presents any stimuli to the user.
CA 02318107 2000-07-20
WO 99!38137 PCT/US99/01415
-36-
As a result, the user is required to actively engage a particular training
exercise
before the training exercise will commence. The described training exercises
therefore wait
for the user to provide undivided attention before continuing with the
training exercise.
Such encourages attention of the user on the training exercise. In addition,
use of each of
the training exercises is limited to a range of time between a minimum
duration, e.g., fifteen
minutes, and a maximum duration, e.g., twenty minutes. The user therefore has
a limited
amount of time to accumulate points. If the user sits idle, the user loses
unrecoverable time
to accumulate points in the token economy toward real, physical rewards as
described
above. Therefore, the user is encouraged strongly to focus and. to sustain
attention to the
training exercises during use thereof.
The nature of some of training exercises 1102-1114 commands undivided and
sustained attention of the user. For example, Old MacDonald's Flying Farm 1102
requires
that the user focus and attentively listen to repetitions of a phoneme to
detect a slightly
different phoneme. If the user loses attention momentarily, the user can miss
the distinct
phoneme and Old MacDonald's Flying Farm notes the incorrect response when the
user
fails to release flying object 102 (Figure 2) within the predetermined amount
of time
following presentation of the distinct phoneme to the user.
To fiuther improve the attention abilities of users, the training exercises
described
above provide frequent set switching, i.e., changing from task to another. In
particular,
each training exercise limits use to the range of times described above.
Therefore, instead
of prolonging use of one training exercise to more fully develop a particular
ability as is
done in conventional teaching systems, trainer 1100 (Figure 81 ) develops a
particular
ability for a limited period of time and switches focus to another ability to
thereby diversify
focus of the user, i. e., to switch the mental set of the user. Such reduces
the cognitive
rigidity of the user.
To fiurther reduce the cognitive rigidity of the user, each training exercise
frequently
switches categories during a session. Specifically, each training exercise
switches
categories upon occurrence of either (i) a predetermined number of responses,
e.g., thirty
(3), for a particular category or {ii) a predetermined proficiency for a
particular category,
e.g., eighty percent (80%) correct responses for any ten (10) contiguous
stimuli. By
limiting any one category to a maximum number of response, fiustration of the
user with
CA 02318107 2000-07-20
WO 99/38137 PCT/US99/01415
-37-
categories with which the user struggles is avoided. By switching categories
upon
achievement of a predetermined degree of proficiency of the user, boredom with
a
particular category and with a particular training exercise is avoided. In
addition,
increasing the frequency with which categories change challenges and improves
the user's
ability to switch mental sets and therefore reduces cognitive rigidity.
These aspects of the training exercises described above significantly improve
behavioral and cognitive abilities of users with such attention disorders as
ADD, ADD/i~,
appositional behavior, conduct disorder, and cognitive rigidity.
The above description is illustrative only and is not limiting. The present
invention
is limited only by the claims which follow.
